Seems promising and I would support these efforts. One danger to watch for and try to address would be attention info-hazards e.g.:
if object-level rejection criteria were specific and public/well-known, they could point less sophisticated bad actors to a curated list of topics/papers/ideas thought by experts to be dangerous
if rejected papers find an alternative publication location, that location could become suffused with a higher density of potentially dangerous material relative to other locations, maybe facilitating malicious application
But these risks can be managed, and are outweighed by the project’s benefits in my view.
Seems promising and I would support these efforts. One danger to watch for and try to address would be attention info-hazards e.g.:
if object-level rejection criteria were specific and public/well-known, they could point less sophisticated bad actors to a curated list of topics/papers/ideas thought by experts to be dangerous
if rejected papers find an alternative publication location, that location could become suffused with a higher density of potentially dangerous material relative to other locations, maybe facilitating malicious application
But these risks can be managed, and are outweighed by the project’s benefits in my view.